1 + 4

rnorm(10)

install.packages("Hmisc")
library(Hmisc)
install.packages("doBy")
library(doBy)

dat_frq  <- read.csv("https://raw.githubusercontent.com/OpenActTextDev/ActuarialRegression/refs/heads/main/CourseCSVData/freMTPL2freq.csv")

str(dat_frq)

dat_frq$X <- NULL
summary(dat_frq)

head(dat_frq)

custom_summary <- function(x) {
  c(
    count = sum(!is.na(x)),
    mean = mean(x, na.rm = TRUE),
    sd = sd(x, na.rm = TRUE),
    min = min(x, na.rm = TRUE),
    `1%` = quantile(x, 0.01, na.rm = TRUE),
    `50%` = quantile(x, 0.5, na.rm = TRUE),
    `99.99%` = quantile(x, 0.9999, na.rm = TRUE),
    max = max(x, na.rm = TRUE)
  )
}

numeric_id <- sapply(dat_frq, is.numeric)
integer_id <- sapply(dat_frq, is.integer)
char_id <- sapply(dat_frq, is.character)
numeric_vars <- dat_frq[,numeric_id]
integer_vars <- dat_frq[,integer_id]
char_vars <- dat_frq[,char_id]
#str(numeric_vars)
#str(integer_vars)
#str(char_vars)
summary_df <- sapply(numeric_vars, custom_summary)
round(t(summary_df), digits = 2);  # Transpose for easier viewing


winsorize <- function(x, p = 0.01) {
  qnt <- quantile(x, probs = c(p, 1 - p), na.rm = TRUE)
  x[x < qnt[1]] <- qnt[1]
  x[x > qnt[2]] <- qnt[2]
  return(x)
  }
dat_frqMod <- dat_frq
dat_frqMod$VehAge<- winsorize(dat_frqMod$VehAge, p = 0.0001)  # like 0.01% and 99.99%
dat_frqMod$BonusMalus<- winsorize(dat_frqMod$BonusMalus, p = 0.0001)
numeric_vars <- sapply(dat_frqMod, is.numeric)
summary_df <- sapply(dat_frqMod[, numeric_vars], custom_summary)
round(t(summary_df), digits = 2);  # Transpose for easier viewing

table(dat_frqMod$ClaimNb)

table(dat_frqMod$VehBrand)
table(dat_frqMod$VehGas)
table(dat_frqMod$Region)
table(dat_frqMod$Area)
table(dat_frqMod$VehPower)

table(dat_frqMod$VehGas, dat_frqMod$ClaimNb)

#install.packages("Hmisc")
#library(Hmisc)
Hmisc::summarize(dat_frqMod$DrivAge, dat_frqMod$ClaimNb, mean)
Hmisc::summarize(dat_frqMod$VehAge, dat_frqMod$ClaimNb, mean)